Welcome to the first edition of Materials Brief — AI tools and workflows for accountants, twice a week, without the hype. This week's news is dominated by AI in audit: a major hallucination incident at EY, two significant platform launches, and a Big 4 firm doubling down on Claude.
1. EY Canada pulls cybersecurity report after AI hallucination investigation GPTZero published an investigation on May 15 finding that 16 of 27 citations in EY Canada's 2025 loyalty systems cybersecurity report were fabricated or broken links — including references attributed to Forbes, McKinsey, Gartner, and WIRED. EY removed the report and said it is reviewing the circumstances. The Financial Times covered it the same day. Going Concern → Read more
2. Optro buys AI-native SOX automation platform Midship Hg-backed Optro (formerly AuditBoard) acquired Midship on May 6. Founded in 2024 and Y Combinator-backed, Midship had raised a $4.15 million seed round and automates controls testing using agentic AI. Midship's three-person founding team joins Optro's product group. Financial terms were not disclosed. CPA Practice Advisor → Read more
3. Grant Thornton launches proprietary AI audit platform firm-wide Grant Thornton US launched gtap (Grant Thornton Analytics & Automation Platform) on May 8, a cloud-based infrastructure that embeds AI agents directly into audit workflows. The platform works with any ERP system and automates workpaper generation and full-population testing. The firm will deploy it to private company audits first, with public company audits following in 2027. CPA Practice Advisor → Read more
4. PwC expands Anthropic alliance to deploy Claude across its workforce PwC announced it will roll out Claude Code and Cowork to US teams first, expanding toward its global workforce of hundreds of thousands of professionals, and will establish a joint Center of Excellence with Anthropic. The firm plans to train and certify a large portion of its staff on Claude. No timeline or cost figures were disclosed. Going Concern (via PwC press release) → Read more
5. Bipartisan Policy Center releases issue brief on GenAI and agentic AI in auditing The Bipartisan Policy Center, with input from the Center for Audit Quality, published an analysis of how generative and agentic AI are reshaping the public company auditor role. The brief examines task-level impacts and workforce implications across the audit function. It is publicly available on BPC's website. Going Concern (via Bipartisan Policy Center) → Read more
6. KPMG survey: AI use in finance functions jumped from 30% to 75% in two years A KPMG global survey of senior finance leaders found 75% are actively using AI in finance functions today, up from 30% two years ago. Within 18 months, 93% of US companies expect to be deploying or scaling AI in their finance functions. Organizations already using agentic AI showed a 32-point advantage over others across six core finance metrics including forecast accuracy, ROI, and close efficiency. Journal of Accountancy → Read more
7. AICPA Auditing Standards Board updates external confirmation standard The ASB approved a new Statement on Auditing Standards on May 14 requiring auditors to use external confirmation procedures for cash held by third parties in most circumstances. The standard also addresses use of intermediaries and direct digital access to third-party data. It is effective December 15, 2028, with early adoption permitted; text to be published in July. CPA Practice Advisor → Read more
Next edition: Tuesday, May 26.